GOES 16 GRB downlink vs GVAR

[GOES 12 GVAR user]

Got another quote for a receiver.
This is the Work Microwave DVB-S2 IP Demodulator.
Models
SDD-IP-L75-BBO $4,550
SDD-IP-V50/L75-BBO $4,950

These are complete stand alone units that are rack mount. One of the nice options is the V50/75 model has a 50 ohm input at a range of 50-180 Mhz so you could use a downconverter at the feed with this unit. Also it can do CCSDS frame output. Quotes are for Base band Frame output.

See spec sheet

However this model will be out of most's price range.
I do think it's worth every penny as what it's capable of doing.

Time to contact Ayecka Comm. systems inc next.

 

[N6BY]

Wow, almost $5K. It does have very high throughput, but GRB doesn't need it since its only 8665K symbols/second.

As for the Ayecka, according to weather01089 his unit sometimes "goes to sleep" for about 1/2 hour or so. But considering the price difference, that problem may be tolerable.

... On Thursday I am expecting to receive a prototype Nooelec saw/filter+LNA combo specially designed for GRB. I think its similar to their Outernet version, but with the filter optimized for the higher frequency. Not sure if the GOES-16 (75W) signal will be strong enough here is California. If not, at least I will be ready for GOES-17 when its up and running at 135W in several months.

Meanwhile I am still trying to get GRBStreamer to lock the GRB signal consistently with TBS cards. But I can get the TBS card to lock NOAAPORT on 89W and stream it.

Anyway, I enjoy technical challenges and this project is definitely a big one!

 

[GOES 12 GVAR user]

Ya, it's expensive but it's got all the bells and whistles, Maindata's is just about as good though at a much lower price which I'm strongly considering at this time.

As for the Ayecka, according to weather01089 his unit sometimes "goes to sleep" for about 1/2 hour or so. But considering the price difference, that problem may be tolerable.

That bothers me a lot, It shouldn't do that. Anyway they have yet to contact me after leaving 2 messages so far. I will not consider them as an option if they won't contact me.

... On Thursday I am expecting to receive a prototype Nooelec saw/filter+LNA combo specially designed for GRB.

Nooelec is the SDR I use to show that SDR spectrum info that I have uploaded here. I have been very happy with it.

Not sure if the GOES-16 (75W) signal will be strong enough here is California

Unless something has changed with the DVBS-2 format used I have a hard time believing that you can't get GOES E there in CA. However if your noise floor is higher then that may be the issue.
Exp. since I read in a earlier discussion that you could get NOAAPORT there.

But I can get the TBS card to lock NOAAPORT on 89W and stream it.

Speaking of (above). Not to get off topic here but I had noticed that NOAAPORT quit working last night (16th) here around 4 AM CST. Have been watching it and finally did a search and found the forum that you guys were discussing of data stream move to Galaxy 89W. That helped me a lot figure out why the feed stopped on SES1. I got no bulletins on the move so it's was quite the surprise. I have been using NOAAPORT to fill the void
without GOES East GVAR. All I use off it is the GOES feed (Pid 102) as the computer I use is not fast enough to sort all the data coming from that link.
Also I want to let you know that there are more channels then 5, there are 8, pids 101-108. 201 is the weather wire feed (text only). If all else fails and you still can't get GOES East then NOAAPORT carries the GOES 16 data. Channel 8 is that feed. But unlike the channel 2 the data is sent as "tilefiles" The images are broken up into tiles. See the GOES R data types image I loaded on the forum under the AWIPS side several pages back (post 84).
Like I talked about several pages back as well the NOAAPORT stream is easier due the MPE format used to find hardware for. Lots of choices for receivers. Wish GRB would have been as easy.

Details of the GOES E channel here
Details of the current settings and receiver here

Now I have to reaim the NOAAPORT antenna, I have had to move all my antennas for this change. GEEZZEE.

Brett, I appreciate the time you and others have put into this and helping me open my eyes on the subject/issues as it's been more then I thought it was going to be. And you're right it's been challenging, though I think there are many watching this forum about this. And I intend to finish putting up my GRB station.

 

[N6BY]

...
Now I have to reaim the NOAAPORT antenna, I have had to move all my antennas for this change. GEEZZEE.

Yah, these NOAA satellites are moving targets over the years. I'm using a motorized 8.5 Birdview for it, and for me it is just a button press from inside the house to move between satellites.

...
Unless something has changed with the DVBS-2 format used I have a hard time believing that you can't get GOES E there in CA. However if your noise floor is higher then that may be the issue. ...

If the new Nooelec device doesn't get me the signal Thursday then I will probably make a new septum feed -- with a circular waveguide like your prototype instead of square.

...
Brett, I appreciate the time you and others have put into this and helping me open my eyes on the subject/issues as it's been more then I thought it was going to be. And you're right it's been challenging, though I think there are many watching this forum about this. And I intend to finish putting up my GRB station.

And I appreciate you starting this thread. I have learned a lot and plan to continue contributing my programming ability to the Open Satellite Project. By summer we should both have working systems.

Hopefully we can get some more people interested and involved in this.

 

[GOES 12 GVAR user]

I'm using a motorized 8.5 Birdview for it, and for me it is just a button press from inside the house to move between satellites.

NOAAPORT has been on SES1 for some time, Aim the dish and leave it been the case here. Having a motorized azimuth adjustment is nice. I have used my eyes and a spectrum analyzer/DVB receiver for the rest. With these feeds coming from geostationary orbit the dish don't have to be moved very often which is nice.

then I will probably make a new septum feed

I would hope not, You put some time and money into that feed. I haven't determined if mine will work as well. Hopefully this Fri. I can find out as the weather is going to warm up for a short bit.
I can suggest this though, You know that the Quaker oats feed worked. So... do what I did
Add another probe and combine the 2 using the methods described below. Heck you could even try to just use the probe. I doubt it will work but you could give it a try.
That's what I have been using here for the tests is the 2 probe method, which gives me about 2 dB of margin of C/N on GRB with a solid lock. I Just don't have a GSE receiver to decode it with yet.
2 equal lengths of short coax to the combiner then to the preamp...done. Combiner has been the cheapest part I have got so far for this project.

And I appreciate you starting this thread. I have learned a lot and plan to continue contributing my programming ability to the Open Satellite Project. By summer we should both have working systems.

Hopefully we can get some more people interested and involved in this.

, That's a good plan a goal. I Agree.

 

[N6BY]

...
I can suggest this though, You know that the Quaker oats feed worked. So... do what I did
Add another probe and combine the 2 using the methods described below. ....

Great suggestion. I will try that before making a new septum feed.

... Made some progress today with weather01089's Digital Devices DVB card. He sent me a GRB stream dump and it is raw BBFrames embedded within a transport stream. I wrote some code to strip away the transport stream and the result was 100% valid raw BBFrames, suitable for ingestion by the GRBDump program. But I have to figure out how to reliably lock the GRB signal. Then we have to see if the Digital Devices card doesn't drop frames now and then like the TBS card.

The Digital Devices card might turn out to be a very low cost GRB receiver.

 

[GOES 12 GVAR user]

Brett,
I just tested my septum feed and it don't work. I can't get a lock even though the signal looks great on the SDR spectrum. It's even a 2-3 dB (-107 dBm) more then what I get with the 2 probe method but it won't lock on either polarization. It will only show a lock for a sec or 2.
The reason for this in my thought is...
1. There could be a linearity to the feed receiving this dual polarization transmission. You mentioned this as stating it could receive linear polarization as well.
If it can receive liner polarization then it would be 3 dB down due to the antenna used. In my test I'm not seeing a 3 dB signal difference.
2. Due to the dual polarization nature of the transmission it could be cross talking between the two resulting in contamination of the received signal wanted.

Since we have both built these one of us should be able to get a semi solid lock on the signal but we can't. So this out rules design flaws as an issue here.
Ever since I built my dual probe feed I have never had any issues getting a signal lock on the DVBS-2 outer layer with the S300. Eventhough I can't process the data layer.
It's possible that this feed may need a bigger dish if it's splitting the signal between the 2 (but I don't know). In any case this won't be doable so we know what works. This is why I held back posting any info on this feed as I didn't know if it was going to work for this purpose.
I consider this a failed test, and will use the 2 probe method for reception of GRB. The 2 probe method will receive 1 polarization of the GRB stream.
What I did learn and gained was using a little larger choke ring does seem to improve the signal.
The current choke ring used is 12" dia X 3 3/8" deep and 1/4" back from the mouth of the feed.
The new choke ring will be 13" dia X 2 5/16" deep (2.32") and 1/4" back for the mouth of the feed.
The measurement is from the outer edge of the choke ring to feed mouth.
The feed I have been using for this is posted below. The dimensions are corrected and proper in this posting.

The simple way to combine the 2 probes is use a 90 degree signal combiner (part Minicircuits ZX10Q-2-25) use a terminator cap and you're done. A cost of around $35-$40 in all. Again use 2 equal lengths of coax to the combiner. Connect the preamp after the combiner and filer if desired (amp again if filter is used) then down the line to the receiver.

Now for the other bit of news.
I finally got ahold of Ayecka. I used their direct email contact. I have been talking with them and will have some more info up here of the outcome of that.
I do have a question though.
Brett, does weather01089 have the SR1-GSE model, can you find out please?

 

[N6BY]

Brett,
I just tested my septum feed and it don't work. I can't get a lock even though the signal looks great on the SDR spectrum. It's even a 2-3 dB (-107 dBm) more then what I get with the 2 probe method but it won't lock on either polarization. It will only show a lock for a sec or 2....

It might not be a problem with your septum feed. I think you need to amplify the signal up to about -70dBM to get a DVB receiver to lock it.

@hdoverobinson from the Open Satellite Group sent me a Nooelec saw filter/LNA optimized for 1680MHz. I placed it about 8 feet down the cable from the Septum feed (near the ground). Then I put a 20 dB LNA at the receiver end and was able to lock the GRB signal last night for the first time. It locked at about 6.2 dB SNR. In GRBStreamer it showed that about 1% of the packets had synchronization errors. But I was able to get some valid data, since I saw the words "Solar Imagery" in a file dump.

...
1. There could be a linearity to the feed receiving this dual polarization transmission. You mentioned this as stating it could receive linear polarization as well.
If it can receive liner polarization then it would be 3 dB down due to the antenna used. In my test I'm not seeing a 3 dB signal difference.
2. Due to the dual polarization nature of the transmission it could be cross talking between the two resulting in contamination of the received signal wanted....

I think its reason #2. I think it was a bad idea for NOAA to combine both RHCP and LHCP at the same frequency. But were stuck with it, and septum feeds show promise.

...
It's possible that this feed may need a bigger dish if it's splitting the signal between the 2 (but I don't know). ..

A bigger dish would help. But I recommend trying the Nooelec LNA/Saw filer for 1680 MHz first, plus another 20 dB LNA. With that combo I got 6.2 dB SNR in California. It should work better farther east where you are.

...
I finally got ahold of Ayecka. I used their direct email contact. I have been talking with them and will have some more info up here of the outcome of that.
I do have a question though.
Brett, does weather01089 have the SR1-GSE model, can you find out please?

I sent him a message asking that. Will let you know what he says.

... Below are some screen shots of the GRBStreamer with the signal locked, and a portion of the data received.

 

[GOES 12 GVAR user]

I think you need to amplify the signal up to about -70dBM to get a DVB receiver to lock it.

No, I can get lock below that. See post #26 where I was using the S300 and showed no signal bar and had lock at 9 dB C/N with about 2 dB margin with the 2 probe feed.

It locked at about 6.2 dB SNR.

A C/N of 6-6.5 is at noise threshold, putting more gain don't help that issue. That issue is made or broke at the first gain stage after the antenna.
The first gain stage should be right after the antenna with a very short jumper. This gain stage sets the NF and overall performance of the system.
You can get some data even though you see errors but it will be hard to get a lot.

If the septum feed is splitting the signal then a 10' dish won't be enough for a solid lock. I will use the method that has worked here for my needs. The septum feed can be looked at again when my 12' dish becomes available.

Thanks for asking him about that receiver.

Keep a lookout for the status files in the GRB stream as they are sent every 5 min. These will be compressed text files.

Good job overall, A little closer each time.

 

[N6BY]

@weather01089 was using an Ayecka "custom firmware test version". I say "was" because I modified GRBStreamer today and now it works flawlessly with his DIgital Devices Cine card. But the Ayecka takes a long time to lock (sometimes an hour) and goes to sleep now and then.

The LNAs are needed to bring the signal up to a level that the TBS card recognizes. Without the extra LNA my TBS card saw nothing (flat line for a signal). But you're right, they don't help SNR.

A 10 foot dish, a septum feed, LNAs, a Digital Devices Cine card, and my GRBStreamer program worked perfectly in Boston, MA today! The $150 Digital Devices card absolutely blows away the more expensive DVB receivers for this purpose.

 

[GOES 12 GVAR user]

was using an Ayecka "custom firmware test version".

But the Ayecka takes a long time to lock (sometimes an hour) and goes to sleep now and then.

WHAT! an hour to lock? I would be unglued for a receiver to take that long to lock.
I am wondering if the firmware/Model has been updated since he got his, what they told me there haven't been any issues with something like that. Hmm... interesting.

The LNAs are needed to bring the signal up to a level that the TBS card recognizes. Without the extra LNA my TBS card saw nothing (flat line for a signal).

I can see that as an issue, the signal we are working with here is weak so I do understand the reasoning.
I do want to make sure that others understand that adding amps in line don't fix a antenna problem.

A 10 foot dish, a septum feed, LNAs, a Digital Devices Cine card, and my GRBStreamer program worked perfectly in Boston, MA today! The $150 Digital Devices card absolutely blows away the more expensive DVB receivers for this purpose.

Well that's great , I said when I started this forum that the DVBS-2 format has made reception of this data possible at a much cheaper cost. It's just how you want to do it. No all receivers are equal as the firmware and electronics make them very different.
I see a CONUS on the RHCP side of GRB there. Now he just needs to keep an eye on the storage drive as it will fill quickly with all the data coming from GRB.
Congrads to both of you to get the job done.

Do post the model of the receiver card so others know.

Welcome to the world of High Resolution Weather Satellite Data.

 

[N6BY]

Thanks. Persistence paid off.

@weather01089 is using a Digital Devices Cine S2 V7A: Twin Tuner TV Karte DVB-S/S2 (PCI Express Karte) - DD Cine S2 V7A

So last night I decided to try to tweak my dish setup here. I wrapped my LNA/Saw filter with aluminum foil and grounded it. It increased the SNR to 6.6 dB.

With that simple fix, I have been streaming for quite a while with 0 errors. Over 1.2 million error free packets here and counting! So I may have written off my TBS6983 too soon.

Now I have to focus on GRBDump because my system is not fast enough to handle all the data. I think I can modify GRBDump so it only processes certain products, not every one.

So it is possible to receive GRB in California with a septum feed and 10 foot dish!

 

[GOES 12 GVAR user]

Thanks for the receiver info.

I wrapped my LNA/Saw filter with aluminum foil and grounded it.

I thought we got that worked out several pages ago, all RF components/blocks should be shielded. Any exposed stripline can receive unwanted noise and signals. The only thing you want to come out of the shield is the connector. It's a must. I know that I told you that active RF parts have to be shielded but so do the striplines. Your RF section is only as good as it's weakest part.
Striplines are transmission lines, but due to the design they are on PC board and not shielded or only partly shielded. When running between blocks of the system you should use a high quality coax such as LMR100A or better if the run is longer. This small coax work great for jumpers I use for going from the feed antennas to the combiner then to the amp.
I currently use LMR240 for the 15' run from the amp the the SDR.

I have been streaming for quite a while with 0 errors. Over 1.2 million error free packets here and counting! So I may have written off my TBS6983 too soon.

So the problem of the TBS card dropping 1 in 200K frames is no longer an issue?

Now I have to focus on GRBDump because my system is not fast enough to handle all the data.

That satellite is pushing some products out. That's just one of the streams.
I agree that GRBDump should be able to be "product selectable". There are many products on the GRB stream that most won't or very seldom use and are for specific purposes.

So it is possible to receive GRB in California with a septum feed and 10 foot dish!

I saw no reason why you couldn't get it there. Though I am still not sold on the septum feed as of yet. Congrads

 

[N6BY]

...
So the problem of the TBS card dropping 1 in 200K frames is no longer an issue?...

The other night the sky was clear and it was relatively cold. It got over 2 million good frames with zero errors before I turned it off.

But during the day today the error rate was about 5%, which is extremely bad.

Right now its dark and raining and the error rate is less than 1%, but still not acceptable.

Everyone in my neighborhood has solar panels (including me). That may account for some of the problem during the day. But it could also be from the 1700 and 1900 cell phone bands, or who knows what else?

Another possibility is that the TBS card might have a sub-optimal implementation of FEC compared to the Digital Devices card? All DVB receivers have to process the low level FEC frames and turn them into BBFrames.

The only way to tell for sure would be for me to buy a Digital Devices card and see how it does under the same conditions. But I'm not sure I want to spend $200 on one, since the card doesn't support blind scan, SNR, and many other features that I would want for general purpose use.

 

[GOES 12 GVAR user]

I have been looking/thinking about the septum feed and due to it's design I'm thinking that it's splitting the signal between both probes (as it should) all the time. The same that when you breathe in/out your nose (when it not plugged up) and the air is divided between the septum.
My original thought was that there wouldn't be a signal reduction unless both probes were used.
This is what have experienced in the past with splitting coax, The signal didn't change until both receivers were put on as a load.
Due to the septum feed waveguide already by design divides the signal (My theory). If this is indeed the case that feed is dividing the 10' dish signal by 2 plus any losses that come with it.
With what I have tested there's not enough signal on a 10' dish to do that. The result is lowering the signal level right to the noise floor.Which would explain the signal reduction from 9 dB C/N (2 probe feed) to no lock condition with that septum feed I tested. That's at least a 2-3 dB in signal reduction.
6-6.5 dB C/N is as low as you can go which you are at Brett. Any slight change in path loss, atmospheric conditions or noise level could cause you to lose lock.
The septum may work good enough in Boston but in your area it's at threshold.
I will recommend that you make yourself a combined 2 probe feed. If my theory is correct you should have a signal level of around 8 dB or better C/N with it which will get you a solid lock on one side.
Unless you know of anybody that is using a 10' dish to get both polarization's, I have never expected to do that with a 10' dish as it would require two 10' dishes or a 12' dish to overcome the C/N factor with GRB.
Again the septum is dividing it weather you are using the other side or not.
That's why I'm sticking with my combined 2 probe feed here as I move forward with this project.

Another possibility is that the TBS card might have a sub-optimal implementation of FEC compared to the Digital Devices card?

Considering that you are right at the noise floor I would consider it to be doing pretty good overall. Though as I stated before another card could help that issue but you are playing a close call with the carrier so close to the noise floor. Could be worse too with another card.

The only way to tell for sure would be for me to buy a Digital Devices card and see how it does under the same conditions. But I'm not sure I want to spend $200 on one, since the card doesn't support blind scan, SNR, and many other features that I would want for general purpose use.

It don't show the C/N level? Wow. How do you know what your signal quality is. Ya, not all support blind scan.
Before you spend that much money on another card try building the 2 probe feed.

 

[N6BY]

No, The Digital Devices card doesnt have C/N level. But I think it does have signal strength and quality information available to the program.
I went ahead and ordered one from Europe. If it only works at night in clear weather for GOES East thats's OK. But when GOES West comes online I should be good 24/7.

None of us in the Open Satellite Group were able to get a decent GRB signal for a DVB receiver without using Nooelec's prototype 'Sawbird' or 'Sawbird+' filter/lna combo. You have to call or email to order one -- they are not listed in their catalog.

... Tonight I am modifying GRBStreamer to have the option of outputting CADU frames for ingestion in CSPP Geo GRB Software. What's unexpected to me is that the CADU frames are not aligned on BBFrame boundaries -- they almost always span them. But the program code to handle that is almost trivial.

 

[GOES 12 GVAR user]

But I think it does have signal strength and quality information available to the program.

I figured it had some kind of signal strength info. That would be strange if it didn't. Even the the cheap china box had that info.
Hope the new card works out for you.

Thanks for the filter info, I do have a tuned bandpass filter for 1687 MHz but I am still trying some different feed ideas here as I think I can get another 2 dB out of my setup for what I learned with the septum feed design. My next step and so far it's been a difficult one is finding a receiver that meets my needs for GRB. Then I'll get into the details of blocks between the antenna and receiver.

... Tonight I am modifying GRBStreamer to have the option of outputting CADU frames for ingestion in CSPP Geo GRB Software.

I just noticed that when you posted that CSPP Geo GRB uses the CADU frames, which is the next frame under the the BBF to ingest the stream. When most receivers output BBF frames, that software uses a post processed BBF's. Am wondering "why did they do that?".
I may know the answer to that question and I'll keep it to myself.
Is there any answers for your trivial challenge found in the code of GRBStreamer? Just asking.
Since GRB streamer has to process those frames somewhere in it's code to generate those images....
Otherwise you're writing part of the decoding software. I say thanks again.

 

[GOES 12 GVAR user]

I played artist with the PUG graphics in (by the numbers) showing what the BBFrames and the CADU's are together.
It would seem, (again by the numbers) that there are
very close 3.5 to ea BBF OR 7 CDAU frames to every 2 BBFrames.
So it would also seem that every 2 BBFrames the CADU frames are aligned?
I'm no programmer and I know this is much harder to write code for than the graphic looks.
Well, anyway I put this together and it may or may not be accurate.

 

[N6BY]

...
I just noticed that when you posted that CSPP Geo GRB uses the CADU frames, which is the next frame under the the BBF to ingest the stream. When most receivers output BBF frames, that software uses a post processed BBF's. Am wondering "why did they do that?".
I may know the answer to that question and I'll keep it to myself.
Is there any answers for your trivial challenge found in the code of GRBStreamer? Just asking.
Since GRB streamer has to process those frames somewhere in it's code to generate those images....
Otherwise you're writing part of the decoding software. I say thanks again.

I can think of 2 possible explanations of why CSPP Geo GRB requires CADU frames:
1) They wanted to reduce the processing load in their program.
2) They wanted to make it so that only select, specialized and very expensive receivers would work with their software.

Nonetheless, I managed to adapt GRBStreamer to give the option of streaming CADU frames. It turned out to be much more difficult than I expected.

I played artist with the PUG graphics in (by the numbers) showing what the BBFrames and the CADU's are together.
It would seem, (again by the numbers) that there are
very close 3.5 to ea BBF OR 7 CDAU frames to every 2 BBFrames.
So it would also seem that every 2 BBFrames the CADU frames are aligned? ...

Its close to 3.5 (7264/2048 = 3.546875). So amost every CADU frame is not aligned to its BBFrame.

... It was clear last night and I got an error free signal for a few hours. I was able to use GRBStreamer and GRBDump to capture a few GRB images.

The images are far from impressive since they were captured at night. But they are some of the first that I was able to get with my 10' dish here in California.

With a homemade septum feed and streaming software that I wrote myself, I'm certain that I now have by far the least expensive GRB receiving system in the world.

 

[N6BY]

Today, @weather01089 was able to run two instances of my GRBStreamer program simultaneously streaming CADU frames, one for RHCP and the other for LHCP. He streamed both signals to CSPP-GEO software and was able to process all 16 channels of the GRB data simultaneously!

As for me, I'm waiting for delivery of my Digital devices card and hoping it cures my problem with marginal reception. If not, I'll have to wait for GOES WEST this summer. With reception of all channels it is possible to combine them and generate natural color images. I really look forward to doing that here soon.